Research
Research Projects
​​
On-the-run Premia, Settlement Fails, and Central Bank Access (WP)
The premium on "on-the-run" Treasuries (i.e. the most recently issued ones) is an anomaly. I explain it using a model in which primary dealers hold inventories of Treasuries. Primary dealers are more likely to hold large inventories of on-the-run Treasuries. There is also less variation across primary dealers in the available stock of on-the-run Treasuries compared with all other, so-called off-the-run Treasuries. Because on-the-run Treasuries are easier to find, they trade at a premium. My theory is consistent with the USD 40 billion of Treasury contracts that fail to settle each day with the median failure rate of off-the-run Treasuries being almost twice that of on-the-run Treasuries.
I use the model to analyze the effects of granting access to central bank facilities to non-banks active in the Treasury market. Broad access stimulates trading and reduces the on-the-run premium, but settlement fails increase and, counterintuitively, only primary dealers benefit.
​
Credit and Anonymity (WP) (submitted, R&R at the Journal of Economic Theory)
with R. Taudien
It is commonly believed that borrowers cannot be anonymous in unsecured credit relations because anonymity heavily reduces the scope for punishment and therefore makes credit unfeasible except for very special circumstances. However, we demonstrate that credit is generally feasible even if borrowers are anonymous. In particular, we construct equilibria where borrowers use potentially multiple pseudonyms (such as usernames or wallet addresses) to interact with lenders. We assume that the complete history of past actions committed by a pseudonym is public but not the identity behind that pseudonym. While borrowers cannot be directly punished due to their anonymity, there is still scope for punishment. One possibility is based on the loss of reputation accumulated by a pseudonym over time. Another involves charging a fee to create pseudonyms. Although credit and anonymity are not mutually exclusive, we also show that maintaining a borrower’s anonymity is costly.
​
Transaction Costs, the Price of Convenience, and the Cross-Section of Safe Asset Returns
with R. Juelsrud, P. Nenov, and O. Syrstad
We study the cross-section of safe asset returns using a tractable asset pricing model with multiple safe assets, agent heterogeneity, transaction costs, and aggregate risk. Changes in the supply or in the transaction costs of a single safe asset induce purifying/polluting effects on the convenience yields of all assets via an aggregate price of convenience. An increase in aggregate risk or risk aversion in our model ends up decreasing liquidity premia via a safety value channel - a repricing effect on risk-free (safe) assets due to a flight to safety adjustment in agents' portfolios.
We test the predictions of our model using data on US Treasury yields and changes in Treasury supply. Consistent with our theory, we show that the convenience yield defined as the difference between a maturity matched Treasury-OIS spread and the 3-months Treasury-OIS spread increases with the supply of long maturity bonds and decreases with the supply of shorter maturity bonds. It also increases with the MOVE Index, which is closely correlated with illiquidity and transaction costs in the Treasury market. However, it decreases with the VIX, consistent with our safety value channel.
Overall, our tractable model can be useful for analyzing the asset pricing effects of central bank market operations as well as unconventional monetary policies.
​
One-sided Market Pressure and Interest Earnings: an Explanation for Covered Interest Rate Parity Deviations
Deviations from covered interest rate parity are a puzzling feature of foreign exchange markets. I present a model of the FX market featuring two components which jointly explain the observed phenomenon - the deviations. The first component is regulatory costs due to one-sided market pressure and the implied accumulation of positions on the providers’ balance sheets. They lead to the observed deviations. After accounting for these balance sheet costs, the parity deviations reflect not an arbitrage opportunity anymore but are costs which were not taken into account. The second component is small interest earnings arising through differences in the commercial paper rate, the Libor rate, and the risk-free investment rate. The interest profits decrease the overall costs.
The model is estimated and tested for the USD/Yen deviations using cross-currency swap data. The fluctuations of the interest earnings are the main driver of the fluctuations of the parity deviations. The measured balance sheet costs are in line with the observed capital costs under the Basel III regulations. In addition, the model results are confirmed on a broad basis by adding the first difference in the interest earnings as an explanatory variable to the multi-currency regression run by Avdjiev et al. (2019).
​
From last Resort to first Resort: Security Lending Facilities in Practice - Evidence from Sweden
with M. Blix Grimaldi and D. Vestin
​
​
Other Projects
Exercise Manual to Macroeconomic Analysis
with D. Niepelt, L. Driussi, and C. Lardy, The MIT Press, 2019
Data Project: Development of a Methodology to estimate the Country Breakdown of Swiss Tourism Exports and Imports
data is regularly published on the SNB data portal (link), 2018
​
The Effects of a Countercyclical Capital Buffer in a DSGE Model for Switzerland
master thesis, 2017
Marktanalyse und Abschätzung der Marktentwicklung von nicht-medizinischen genetischen Untersuchungen
translated English title: Market Analysis and Estimation of the Market Development of non-medical genetic Testing
with M. Frey on behalf of the FOPH, 2015
​